172
Jouml of Natural product^ Vol. 58, No. 2, pp. 172-1 76, February I995
IN VIVO AND IN VITRO BIOSYNTHESIS OF SAPONINS IN SEA CUCUMBERS RUSSELL G. KERR* and ZHENGJMN CHEN
Department of Chemistry, Florida Atlantic Uniwsity, Boca Raton, Florida 3343 I AesTmcr.-The triterpene precursor of saponins in sea cucumbers has been identified as parkeol [lanost-9(1 l)-en-3P-ol] [l].Dissection of the sea cucumbers Holotburza fiidecl and Actinopyga agassize after incubations with radiolabeled parkeol demonstrated that saponin biosynthesis occurs exclusively in the Cuvier gland. This result was corroborated by incubating a cell-free extract of the Cuvier gland with labeled parkeol and observing transformation of the precursor to saponins.
Saponins are common metabolites of members of the phylum Echinodermata, particularly the seacucumbers(Holothuroidea) and starfish (Asteroidea)(1).Holothurins, saponins from holothurians, are typically composed of carbohydrate and triterpenoid moieties, while asterosaponins are characteristically steroidal glycosides (1).In addition to the echinoderms, sponges, and coelenterates have also furnished a number of novel saponins (2). Most saponins exhibit potent hemolytic properties (3), while antitumor (4), antiinflammatory ( 5 ) , and antibacterial (6) activities have also been reported. The saponins from severalstarfish have been shown to inhibit influenzavirus multireplication (7). Recently, several triterpenoid glycoside sulfates have been isolated from the sea cucumber Cucumaria ecbinata (Cucumariidae) and the corresponding aglycones showed activity against L-1210 and KB cells (8). Due to the general toxicity of the saponins it is generally believed that these compounds act as chemical defense agents. The toxins are concentrated in specialized organs known as Cuvier glands which can be eviscerated to deter predators. The triterpene portions of holothurins typically contain a A'11) double bond, which suggests that these compounds are derived from parkeol 111 {lanost-9(1l)-en-3P-ol] (Scheme 1).A few reports have appeared in the literature concerning the metabolic origin of these echinoderm toxins, yet their biosynthetic precursors remain to be identified. Recently, Cordeiro and Djerassi reported that 3H-labeled parkeol, lanosterol, and cycloartenol were not transformed into the saponins of the sea cucumbers Bobadscbia argus and Holotburia mexicana (9,lO). As these authors have indicated, it is puzzling that none of these triterpenes was transformed into saponins and this result may be due to some factor such as lack of transportation ofprecursor to the site of saponin biosynthesis, or the fact that the toxin is not continuously synthesized. Previous experiments with '*Clabeled mevalonate (11) and acetate (12) resulted in low levels of incorporation into the crude saponins suggesting de novo biosynthesis does operate in sea cucumbers. The sterols of sea cucumbers have been well characterized ( l ) , and generally, the principal sterols have a A'11) l4a-methyl nucleus and thus appear to be metabolically related to the saponins. The biosynthetic precursor of this unusual sterol class has been demonstrated to be parkeol(9). Further, it was shown that squalene cyclizes directly to produce parkeol and not the isomeric lanosterol or cycloartenol. It thus seems reasonable that in sea cucumbers, parkeol E l } is the sole cyclization product of squalene 127 and that 1 is the precursor of both holothurins and sterols. RESULTS AND DISCUSSION The sea cucumbers used in this study were Holotburiafiridea and Actinopyga agassize.
February 19951
Kerr and Chen: Marine Saponin Biosynthesis
OH
173
OH
SCHEME 1. Biosynthesis of saponins in sea cucumbers.
These were collected off Long Key, Florida, and maintained in artificial tide pools with running sea water at the Keys Marine Laboratory. We have found that sea cucumbers survive for many weeks in this environment; however, if placed in relatively small aquaria, they generally eviscerate their Cuvier gland within a few hours. The organisms were left for about two weeks in the tide pools prior to incorporation experiments. It was our intent to provide as natural an environment as possible for the test sea cucumbers, to minimize stress and thus avoid evisceration which would result in loss of any labeled saponins. The saponins from these sea cucumbers were isolated as previously described (13) and identified by comparison of their I3C-nmrspectra with literature data (13).The main saponin from both H. frorideu and A. agassize is holothurin A {3] (13). In vivo biosynthetic experiments were conducted by injecting a solution of {24-3H} parkeol in EtOH-H,O (1: 1) to the coelemic cavity of the sea cucumber. The labeled parkeol was synthesized as described previously (9). The injections were performed underwater to minimize stressing of the animal. The sea cucumbers were monitored for the first few hours to determine ifthe procedure would result in evisceration ofthe Cuvier gland. After an incubation period of seven days, the organisms were harvested and stored at -20'. To localize the site of saponin biosynthesis, the thawed sea cucumber was dissected into three components: (a) body wall, (b) Cuvier gland, and (c) coelemic fluid and internal organs. The three components were extracted separately by Soxhlet extraction with 95% EtOH. Purification of the saponins and parkeol was achieved by prep. tlc followed by repeated injection on reversed-phase hplc. The results of the parkeol feeding experiments with H. floridea and A. agassize are summarized in Table 1. With both sea cucumbers, radioactive parkeol and holothurin A 131 were recovered from the Cuvier gland, while with the body wall and coelemic fluid, the only radioactive fractions
174
Wol. 58, No. 2
Journal of Natural Products TABLE 1. In Vivo Biosynthetic Experiments.
I
Recovered radioactivityfrom the injection of 22X106 parkeol El] (in dpm)
Sea cucumber pan
Holothuria f i i d e a
3
1 Body wall. . . . . . . Cuvier gland . . . . Coelemic fluid
...
Actinopyga agassize
i . 6 X 10’
cold
cold
1.9X10’
1.7X102
cold
3
1 6.7 x 10’
cold
1.6XlO’’
1.3X 10’’
”The Cuvier gland could not be separated from the coelemic fluid and internal organs in A . agassize and were therefore worked up together.
were those associated with parkeol or sterols. These data suggest that parkeol is transported to the Cuvier gland and, in this organ, is then elaborated to the saponin. To confirm that the location of saponin biosynthesis in sea cucumbers is the Cuvier gland, cell-free extracts of this organ from H. fIwidea and A. agassize were prepared and used in incubation experiments with labeled parkeol. The enzyme preparation was generated by grinding a fresh Cuvier gland in liquid N, to a fine powder in a large mortar and pestle and adding this to a buffer containing a mixture of protease inhibitors and protein stabilizers (phenyl methyl sulfonyl fluoride, Leupeptin, Pepstatin A, ethylene diamine tetraacetic acid, and dithiothreitol). Two sets of experiments with different buffers were used, as our experiencewith other systems has been that the choice ofbuffer greatly affects the activity of the cell-free extract. The buffers used were Tris and a phosphate buffer both at a pH of 7.7.The in vitro metabolic experiments were performed by incubating the cell-free extract with 1.1X lo6 dpm [24-3H7parkeol at 27’ for 6 h. Controls with both buffers were conducted in which the cell-free extracts were heated at 100”for 1 h prior to incubation with labeled parkeol. The results of all in vitro biosynthetic experiments are summarized in Table 2. Although the level of transformation varied between the two sea cucumbers, radioactive holothurin A f37 was obtained from both cell-free extracts. The recovered saponin from the controls was not radioactive in both cases. With all in vivo and in vitro biosynthetic experiments, the intact saponin was purified by prep. tlc, followed by reversed-phase hplc using MeOH as mobile phase. The saponin fraction from the hplc injection was reinjected using the same conditions, and a small portion analyzed by a scintillation counter. To confirm that the observed radioactivity was due to holothurin A rather than an impurity that co-eluted on tlc and hplc, the purified saponin was hydrolyzed with aqueous HI and the resulting triterpene TABLE 2. In Vitro Biosynthetic Exmriments. Recovered radioactivity from the incubation of 1.1X l o 6parkeol El] with cell-free extract (in dpm)
Holothuria
Type of buffer 1
Phosphate buffer . Control . . . . . . . . Tris buffer . . . . . . Control . . . . . . . .
4.4X10’ 1.3X10’ 4.4X10’ 1.3X106
Actinopyga agassize
3 4.6X 10’
cold 5.4X10’
cold
1 6.6X l o 5 1.2X1O6 9.8X 10’ 1.2X1O6
3 2.8X10’
cold 1.2~10’
cold
February 19951
Kerr and Chen:
Marine Saponin Biosynthesis
175
portion purified by prep. tlc. The recovered radioactivity of the resulting aglycone (H. pwidea, 1900 dpm; A, ugussize, 1300 dpm) was very similar to that of the portion of the saponin used in the hydrolysis (2000 dpm in each case), indicating that the observed radioactivity was due to the saponin. The results obtained in this investigation have identified the triterpene precursor of saponins as parkeol 111 and, through dissection of sea cucumbers following in vivo experiments and by examining the metabolic capacity of a cell-free extract of Cuvier glands, the site of saponin biosynthesis has been established. EXPERIMENTAL GENERAL EXPERIMENTAL PROCEDURES.-H~IC was performed using a refractive index detector, aid either an Ultra-Si normal-phase column( 10 mm i.d. X 25 cm) using 8%EtOAcihexanes, or areversed-phase Ultrasphere column (10 mm i.d.X25 cm) with aqueous MeOH as mobile phase. Prep. tlc was performed with precoated Si plates (1000 p m thickness) using berberine chloride as the spray reagent. All solvents were distilled prior to use. Radioactivity was determined using a liquid scintillation counter and a toluene-based scintillation fluid. ANIMALMATERIAL.-H. fiidpa and A. agassize were collected off Long Key, Florida, at depths of 12 m. Voucher specimens have been kept in the author's laboratory at Florida Atlantic University. They were maintained at the Keys Marine Laboratory (Long Key) in running sea water prior to transporting to Florida Atlantic University. WHOLE ORGANISM BIOSYNTHETIC EXPERIMENTS-whole organism biosynthetic experiments were performed by injecting an EtOH solution of radiolabeled parkeol into the coelemic cavity of the sea cucumber and the organism left for 7 days, flash frozen in liquid N,, and stored at -20".
EXTRACTION OF SEA CUCUh€BERS.-The thawed sea cucumber (ca. 500 g) was dissected into its (a) body wall, (b) Cuvier gland, and (c) coelemic fluid and internal organs. Each of these three portions was extracted separately in a Soxhlet apparatus with 200 ml95% EtOH for 48 h.
PURIFICATION OF TRITERPENES AND SAPONINS.-Parkeol(5 mg) was added as cold carrier to the crude extract as it is only present in trace quantities in the sea cucumbers. The EtOH extract was concentrated in vacuo and partitioned between Et,O and H,O. The organic fraction was shown to contain the triterpene parkeol 111 by tlc and was purified by prep. tlc using precoated silica plates (1000 prn thickness) and a mobile phase of hexanes-Et,O (1: 1). Subsequent purification by normal-phase hplc (8% EtOAcihexanes) followed by reversed-phase hplc (MeOH) resulted in aparkeol fraction that was shown to be pure by capillary gc. The aqueous fraction was concentrated by azeotropic distillation with C,H, and purified by prep. tlc (silica, 1000 p m thickness) using CHC1,-MeOH-H,O (65:35: lo), followed by hplc purification under reversed-phase conditions using MeOH-H,O (73:27) to yield holothurin A [3]. Hydrolysis ofsaponin.-A portion of the hplc-purified saponin (holuthurin A, 3) (2.0X 10' dpm) was hydrolyzed with 57% HI at 120-130' for 6 h. Following an aqueous workup, the resulting genin was purified by prep. tlc (CHC13-MeOH-H,0, 65:35:10). PREPARATION OF CELL-FREE EXTucr.-The Cuvier glands of specimens of H. f i i d p a and A. agassize were carefully and quickly dissected at O", and ground to fine powders in liquid N,. These were added to two separate buffer systems, buffer A, a phosphate buffer, and buffer B, TESiTris, both at a p H of 7.7. The buffers (100 ml) contained 5 mM dithiothreitol, 5 mM ethylenediamine tetraacetic acid, and 10% bovine serum albumin. Leupeptin (100 pg), pepstatin A (100 pg), and phenyl methyl sulfonyl fluoride (0.1 mM) were added to the buffered crude enzyme preparations, followed by Tween 20 (100 PI). Aliquots (5-ml) of the cell-free extracts were then incubated with 0.5 pCi [24-14C] parkeol for 6 h at 27'. Controls of the cell-free extracts were prepared by heating the enzyme preparation at 100' for 1h prior to the 6-h incubation. The incubations were terminated by addition of 5 ml of EtOAc. The parkeol and saponin were purified chromatographically as described above. PREPARATIONOF RhDIOLABELEDPARKEOL.-[24-'H] ing to a previously reported procedure (9).
Parkeol was prepared from cyclomenol accord-
ACKNOWLEDGMENTS Financial assistance was provided by the Donors of The Petroleum Research Fund, administered by
Journal of Natural Products
176
mol. 58, No. 2
the American Chemical Society, and by a Bristol-Myen Squibb Company Award of Research Corporation. We acknowledge the generous assistance of the staff of the Keys Marine Laboratory, Layton, Florida, in the collection and maintenance of the sea cucumbers. LITERATURE CITED 1. 2. 3. 4.
5. 6. 7. 8. 9. 10. 11. 12. 13.
D.J. Bumelland J.W. ApSimon, in: “MarineNatural Products.”Ed. byP.J. Scheuer,Academic Press, New York, 1983, Vol. 5, Chap. 6. D.J. Faulkner, Nat. Prod. Rep., 9, 323 (1992), and previous reviews in this series. Y .Hashimoto, “Marine Toxins and Other Bioactive Marine Metabolites,”Japan ScientificSocieties Press, Tokyo, 1979, pp. 268-288. G.R. Pettit, J.F. Day, J.L. Hartwell, and H.B. Wood, Nature, 227,962 (1970). L.A. Goldsmith and G.P. Carlson,in: “Food and Drugs from the Sea. Proceedings 1974.”Ed. by H.H. Webber and G.D. Ruggieri, Marine Technology Society, Washington, DC,1976, pp. 354-365. G.D. Ruggieri, R.F. Nigrelli, in: “Bioactive Compounds from the Sea.” Ed. by H. Humm and C. Lane, Marcel Dekker, New York, 1974, pp. 183-195. Y. Shimizu, ExpPn’entia, 27, 1188 (1971). T. Miyamoto, K. Togawa, R. Higushi, T. Komori, and T. S d i , Liebigs Ann. Chem., 453 (1990). M.L. Cordeiro and C. Djerassi,J. Org. Chem., 55, 2806 (1990). M.L. Cordiero, R.G. Kerr, and C. Djerassi, Tetrahedron Ltt., 29, 2159 (1988). S.S.N. Murrhy and A. Der Marderosian, in: “Food and Drugs from the Sea. Proceedings.” Ed. by R. Worthen, Marine Technology Society, Washington, DC,1973, p. 181. A. Kelecom, D. Daloze, and B. Tuach, Experimtia, 32, 2353 (1976). G.B. Elyakov, T.A. Kuznetsova, V.A. Stonik, and R. Albores, Comp.Biochem. Phyriol., 52B, 413 (1975).
Reseiwd 5 July 1994